Bicycle detection is one area we find requires further explanation. In 2017, the UK government published the Cycling and Walking Investment Strategy.  One of the targets from this strategy is to double the cycling activity by 2025 to 1.6 billion ‘stages’ or journeys. An honourable and ambitious target.  They also want to reduce the rate of cyclists killed or seriously injured on our roads, and they want to make cycling or walking the 1st choice for short journeys.  Here’s my problem. I haven’t really ridden my bicycle on the road since my teenage years; I just don’t feel safe. Cars are a lot bigger these days, so there is less space for us to share. I also feel that drivers do not appreciate how vulnerable you are when cycling. So I put my bike on the roof of my car, drive to a cycle path and use that.  If I want to increase my cycling in line with government’s strategy, I need to leave the car at home, not put the bicycle on the roof and ride to the shops.

Traffic designers are putting in cycle lanes, segregating them for the traffic, and creating green routes. This is all great stuff but for me to get to my local Co-Op there isn’t the space to put cycle lanes in.  I have to ride down a narrow main road, across two sets of traffic lights, and a great big roundabout.  The traffic lights have a lot of left turners, and they don’t seem to notice that I am there. By profession, I am a Traffic Signal Engineer and I’ve been working in the industry since I was at college, in the late 1980s.  I worked for Ferranti (remember them?) and worked on UTC / SCOOT systems and traffic signal intersections.  In all that time our industry hasn’t really thought about cycling detection.  We put an Induction Loop in the road and hope that they ride over it in the right place and that the Loop is sensitive enough to detect the bicycle.  Not great if the bike is carbon fibre, and I’ve noticed a lot of cyclist don’t ride in straight lines, in the right part of the carriageway and at the correct speed as they approach the stop line, so we don’t always detect them.  And, when we do detect bicycles, we treat them as if they are a normal car.

 

For over 30 years, Induction Loops have been the first choice for traffic signal designers. Mainly because there haven’t been any alternatives and they are good at detecting big lumps of metal, but they are not great at detecting cyclists. So, it has been difficult to innovate with the traffic signal designs, to improve safety and the experience for the cyclist on the carriageway.  In recent years however, new technologies have emerged that have the potential to replace Loop technology. These new technologies can be utilised to detect cyclist and I will now review these based on my own personal experience.

Radar

Radar has been around for years on traffic signals and is mainly used on pedestrian crossings.  The uptake on using radar on junctions in England and Wales is low, although in Scotland it is far more common. Historically, radar had to have a speed threshold to prevent false detection of pedestrians, cats, dogs etc. on the approach to the stop line. They were also unable to detect stopped vehicles, so you need to install a Stop Line Detector, or a Loop. Recent detectors like C&T Technologies’ TMA-122can detect very slow-moving vehicles, but filter out pedestrians, and detect stopped vehicles. They also are very good a picking up cyclists. So, if a carbon fibre bike approaches with the latest radar you don’t need to worry about it.  The biggest enemy of radar is occlusion, where large vehicles block the field of view.  You need to think carefully about where it is going to be installed, and make sure they are not obstructed by street furniture, trees etc.

Magnetometers

These are typically installed in the carriageway. These are used mainly for junctions and are good for MOVA (Microprocessor Optimised Vehicle Actuation) sites. You must be careful on the location of the installation to make sure you can detect them. They do have a battery, so at some point this will need to be changed so there is a maintenance cost.  There is a debate on how good these are for detecting non-ferrous bicycles, and there are versions with built in radar that is designed to detect cyclists. Correct siting of these is important if you get it wrong there may be expensive traffic management costs to correct the problem.

Induction Loops

Loops are essentially a coil of wire that is laid in a slot in the carriageway. While these are very good at detection of horizontal sheets of metal, they are not soot good at detecting vertical sheets of metal like bicycles. To help with this you can install different shapes of Induction Loops that are designed to be used specifically  for cycle detection. From my own experience, these special loops are variable in performance, and conventional XYZ Loops or MOVA Loops are even worse as sometimes cyclist completely miss the -Loop. While you can increase sensitivity of the Loop to help detect cyclists, this can cause other issues, such as false detection from vehicles in other lanes or even vehicles travelling in the opposite direction.

Video Detection

Video detection in the UK currently is not widely used. Around 20 years ago, companies brought video detection in from the USA with little success. They were affected by weather factors such as glare, rain, snow, shadows, wind a typical Summer’s day in the UK!  Technology has advanced massively, and, in the USA, this type of detection is widespread, with the old problems having been addressed years ago. Lenses are self-cleaning and heated, so you don’t need to worry about maintenance once they are installed at height. 

 

There are some big advantages with video detection; you can draw zones, so it is flexible with no traffic management costs to change things. The cameras are also multifunctional or can have normal vehicle detection, detect pedestrians, count vehicles and pedestrians, carry out classification studies, live stream video, and even host data to give you traffic survey data.  The latest generation can even send an output when it detects a cyclist, and an output to illuminate a sign to tell the cyclist they have been detected. (once the DfT allow it -we are working on that!).

Summary of options:

If you just want reliable detection of a cyclist, but you are not looking for bicycle differentiation consider the latest Radar products. These can be used in mixed or dedicated cycle lanes as well as on Compact MOVA. Consider Magnetometers if the cyclist is segregated with dedicated cycling lanes that cyclist use. I would personally try and avoid Loops for cyclists.  Look at Video Detection if you can install the camera without occlusion problems. These can be used in mixed traffic conditions, and on dedicated cycle lanes. Height-wise, a rule of thumb is 1 in 10. So if you want to detect 80m you need to be 8m up.  There are brackets available that will get the camera up about 2m above the height of a traffic signal pole or mount the camera further up the carriageway. Remember the cameras can pick out cyclist in mixed traffic, so you can have extended intergreens, or early start to improve safety.

If you would like to discuss any of these technologies or ideas please feel free to contact me, and I will be happy to discuss your specific sites to find the best solution.